Subsidies and tariffs have very specific meanings. They get at the same goal of increasing domestic production, but one increases government spending while the other increases government revenue through an import tax. Ultimately the domestic consumer pays either through taxes or increased prices, but they are very different mechanisms.
Its overly simplistic to assume that a subsidy will increase prices and harm the economy. We have been subsidizing corn for decades and it has kept prices artificially low. Prices for corn and downstream products like beef are artificially low specifically because of federal subsidies.
If a subsidy creates domestic jobs it could have a net benefit. The ultimate question is what metrics are impacted and at what rate. If domestic jobs are created fast enough to increase demand for employees, wages could rise faster than prices.
I have no idea what the end result would be, but that's ultimately my point. Economic models are just that, models. Models are based on analysis of past events and assumptions if the future. We don't know what we don't know, and we often don't recognize how inaccurate models are when we simplify them down to a handful of measurements.
To your original point, hospitals do seem reasonable to focus on with regards to stability through simplified supply chains. Chips may help there, our hospitals are dependent on a huge amount of tech today. Paper, silicon, and metal products would also likely be high on the list as they are often treated as consumables - the consumable nature of those products may make them more important than chips.